HDR Waterfall

[duncan22]

Can someone clarify the point made about only one photo being needed for HDR ?
I Know you can take one RAW photo and then make 2 further copies one at +EV and one at-EV using photoshop or a number of other programs but there must be a limit on how much you can push them compared to taking 3 RAW photos and processing them in photomatix or other similar program.
To put the question another way would you see a clear difference in a A4 print of a HDR image where both of the above methods were used to produce the image ?
If the answer is no then why do people use Photomatix and other HDR software ?

[Forseti]

Well I'm certainly no expert on HDR although I've read a bit it. With that caveat here's my view on the subject which is of course open to correction should the real experts in HDR pop along.

First and foremost, as in any area of photography, it's important to understand exactly what is meant by 'dynamic range' and here's a very detailed explanation if you have the mind to read it. http://www.cambridgeincolour.com/tut...amic-range.htm

Even prior to reading that article I found myself in disagreement somewhat with Gil Richie's comment where he states that '......just give the image different exposures - you don't require multiple images....' The reason for my disagreement is that the sensor on your camera is only capable of capturing a pre-determined dynamic range and that is a physical characteristic. By dialling in plus and minus exposure you are simply upping or lowering the points at which the dynamic range will be captured, not the overall range. If you imagine using a 12" ruler to measure a 24" table top you would have to move the ruler sideways a complete length i.e. 2 x 12" to measure the table - however you are still only using a 12" ruler which is the physical characteristic and is fixed.

If you take a captured image into Photoshop with a certain captured dynamic range you can of course increase/decrease exposure - in fact you can go so far as to see nothing but either a black or white screen. All this of course will be at the expense of other captured data. You can see this for yourself by opening up a RAW image in Photoshop and drastically up the exposure. When doing this observe the histogram and in particular the pixel densities in all the channels. A bit like the ruler in some ways - if it were made of rubber you could stretch the 12 inches out to make 24 inches but the probable consequence would be that the ruler would decrease in breadth (flatten out). So in short, what you gain on the roundabouts you lose on the swings.

To this end I believe what Chris (I think it was) who said that a single RAW image that is processed pseudo HDR cannot be considered true HDR. Just my view on the subject anyway.

[dpower]

x

[nickanderson]

Well maybe I will visit yours instead of mine in that case

[dpower]

x

[chris-p]

Originally Posted by dpower

I use one raw image for subjects that have a lot of motion for the HDR process. The image is processed with the camera's raw program, Nikon NX2 to make images with + and - EV. Even if the image is not processed to produce an HDR, the full range of a single raw image will not be displayed on a computer monitor. The + and - EV just shifts the "visable window" that can be viewed. Tone compression makes the image seem to have greater DR. When the images are first processed into an HDR they are very ugly without tone compression. Three (or more) shots of moving subjects gives results that are not what I want. Changing the raw EV gives more DR ... the end results look the same, no matter what I name you give them. The objective is to get details in the shadows and highlights.
I like to have fun making phots, hoping others also enjoy them, rather than massaging the linquistics of the process.

I understand why people under and over expose a single RAW file, especially when it comes to movement in the images but the statement...

Originally Posted by dpower

Changing the raw EV gives more DR ... the end results look the same, no matter what I name you give them

is plain wrong. It simply defies the laws of physics. It cannot be. You can't create dynamic range out of thin air. The dynamic range of the sensor is a fixed thing and every shot you take at specific exposure settings has a maximum and a minimum exposure value.
Oh, and the end results don't look the same - you said that yourself! Using more than one shot can introduce movement and overlaying the same RAW file amplifies noise. Both of which are different. And that's ignoring the difference in actual DR levels.

Altering the EV later on can't extend the dynamic range, it just makes the image brighter or darker. You can combine images, just like with an HDR image but they're not "high dynamic range" they're pseudo-HDR or DRI (take your pick).

I should also add that I'm not criticising the method, the photographer or the image. Given the subject, there is no other way to capture it, and as I said above its well composed and processed.

Originally Posted by duncan22

To put the question another way would you see a clear difference in a A4 print of a HDR image where both of the above methods were used to produce the image ?
If the answer is no then why do people use Photomatix and other HDR software ?

Depending on the individual shot the answer is usually yes. Overlaying the same RAW file onto itself has the advantage of their being no movement or frame shift but you also amplify things like noise. This isn't necessarily a problem but it depends on how far you've over exposed the file and what your ISO was etc. Suffice to say, it can make noise a lot worse.
DRI is usually sharper, again mainly because you don't have to worry about subject movement and frame shifting which lead to smudging of fine detail.

[Gil Ritchie]

Disagree with me "somewhat" if you like forseti - but you are wrong - I was extending the dynamic ranges of images - long before the expression HDR - became "fashionable"

The dynamic range of an image - in raw format - can be extended at both ends (highlights and shadows) by simply increasing the exposure in one case and decreasing the exposure in the other case - then merging the resultant images - the position of the pixels are NOT altered in anyway - only their luminosity will be altered - giving an true HDR image.

If,however, you use the so called "proper" HDR method, that most use - ie using 3 or whatever number of separate images - you will also extend the dynamic range substantially - but the pixels will NOT be the same - as they are from difference images - this can be somewhat problematic when you have movement in landscape images (clouds,water,trees in the wind etc etc) - still life - does not have these problems - but irrespective of this - when you use these multiple images - you are in fact creating a composite HDR image and not an out of the camera single HDR image.

The expression "pseudo HDR" is wronly applied to the former method - as using one RAW image in such a fashion produces a real HDR effect - the word "pseudo" only applies if you were to use JPEG - rather than RAW.

I,thus, fully stand by my original posting regarding this matter.

Having said all that - I fully agree with the dpower's (the original poster) sentiments - and I will also leave any further "massaging of linguistics" to those that spend most of their time engaged in such "trivia" - like him - I enjoy the viewing and creating of images - rather than wasting time in such meaningless arguements.

As the original posting in this thread had plenty of movement - the proper method was that used by the poster - and used well - a truely great HDR image.

[Forseti]

Originally Posted by Gil Ritchie

I was extending the dynamic ranges of images - long before the expression HDR - became "fashionable"

Well of course Gil I took that as fact as in all probability you were gathering wood for your Ark long before Noah was even born.

As regard movement - a couple of interesting DTown TV videos here http://kelbytv.com/dtowntv/ . Episode 25, (14th Jan) includes a segment where the capture was made etc and the bonus episode of the same date shows how the HDR was processed. And yes, it's all about movement. That said, it's presented by Matt Kloskowski who only started on the HDR after it became fashionable so some may ask 'what does he know?.